NEW FRONTIERS OF THERMODYNAMICS. The science of heat flow might have begun in the 19th century, as a way of maximizing engine efficiency, but thermodynamics is still a forefront discipline, especially when it attempts to maximize information flow in computers. Typical of the new work is a pair of articles which arrive at some surprising conclusions.

In the first paper, Armen Allahverdyan of, CEA Saclay (France)/University of Amsterdam (Netherlands)/Yerevan Physics Institute (Armenia), aarmen@spht.saclay.cea.fr, and Theo Nieuwenhuizen of the University of Amsterdam (nieuwenh@wins.uva.nl, 011-31-20-525-6332) state that there may be a new, previously overlooked work-producing process in classical heat engines. All heat engines are driven by a temperature difference between two reservoirs or "baths." Usually the hot and cold reservoirs are isolated from each other--they interact via an intermediary, namely the "working substance" such as a gas which receives heat from the hot bath, pushes a piston, and sends unused heat to the cold bath.

Now, the researchers suggest that putting the hot and cold baths in direct contact for relatively short amounts of time may result in useful work, if the baths have very different relaxation times, the amount of time that each takes individually to come to thermal equilibrium. One could accomplish this, for example, by putting together a very large bath and a very small bath. Traditional treatments of thermodynamics assume that direct interactions between the hot and cold reservoirs result in the dissipation of energy and nothing else. But these treatments make the simplifying assumption that the baths interact very feebly or for infinitely long times.

The authors show that allowing the two reservoirs to interact for short windows of time (relative to the time in which two reservoirs arrive at an equal temperature) may bring about a transfer of energy that can be converted to work, in cases where the final common temperature of the baths is lowered by their direct interaction. If verified experimentally, this proposal may lead to new engine designs and perhaps revise estimates of the maximum efficiency of a heat engine. (Allahverdyan and Nieuwenhuizen, Physical Review Letters, 10 July /pnu/2000/.)

In the second paper, the same research team suggests that a quantum particle (such as an electron) interacting strongly with a reservoir of particles may violate the Clausius inequality--one formulation of the second law of thermodynamics, which states that it is impossible to do work without losing heat. What the researchers term "appalling behavior" can be traced to the quantum mechanical property of entanglement, in which a quantum particle (such as an electron) is so strongly interlinked with another particle or group of particles that the resulting behavior cannot be treated by standard thermodynamic approaches.

In this paper, the Amsterdam scientists study the entanglement of a particle with a "quantum thermal bath," a reservoir of particles with which the first particle can exchange energy and momentum. According to the researchers, entanglement prevents the quantum bath from observing the normal requirements for a heat bath. Therefore, thermodynamics simply cannot say anything useful about the system.

Standard thermodynamics dictates that the bath be in thermal equilibrium and not interact strongly with an external object. To the contrary, the bath strongly interacts with something external to it (the entangled particle) and it cannot reach equilibrium, since it constantly exchanges energy and momentum with the particle. At low temperatures where entanglement could be easily preserved, the researchers state that this system can apparently violate the Clausius inequality--in which the heat gained by the particle must be less than or equal to the temperature multiplied by the change in its entropy (or disorder). Near absolute zero temperatures, a situation which would ordinarily require the particle to lose heat, the researchers show that the particle could gain heat, by the Clausius relation.

According to this scenario, applying a cyclic parameter such a periodically varying external magnetic field can cause the entangled particle to extract work from the bath--something forbidden in a classical system. Further, the researchers say that this phenomenon could be said to constitute a perpetual motion machine of the second kind. However, they are quick to point out that the particle can only extract reasonable--but not limitless--amounts of work, as the bath must maintain a minimum ground-state energy rather than be completely exhausted. (Second paper: Phys. Rev. Lett. 28 August /pnu/2000/; Select Articles.)